Search

Zero-dependency TypeScript security filter for LLM streams — redact secrets, enforce tool policy, byte and event modes.

LLM Router local routing proxy for OpenAI-compatible Chat Completions APIs.

Cross-agent SDD automated development workflow kit (ADI/ADI-like).

Batch switch ChatLuna conversation models for Koishi

Cawdex — terminal coding agents with a mind for the whole repo. Speaks any OpenAI-compatible API with repo-aware context, benchmark tracing, MemPalace memory, and terminal-first agent workflows.

Codex CLI adapter for [`@lythos/agent-adapter`](https://github.com/lythos-labs/lythoskill/tree/main/packages/lythoskill-agent-adapter/). Spawns `@openai/codex` agent processes for arena cross-player mode.

Skill mestre — guia funcionários não-técnicos da meutudo a criar, empacotar e publicar a própria skill Claude Code no npm, com auto-update e reporting de erros via Slack injetados automaticamente

A confidence-calibration skill for any AI coding agent. Tags every claim 🟢 [VERIFIED] / 🟡 [INFERRED ~80%] / 🟠 [GUESSED ~30%] / 🔴 [UNKNOWN] so you can scan how much to trust each line.

Terminal-based developer chat assistant. Permanent hard fork of gsd-pi with LangFlow flow triggers, a flow builder, and optional gateway routing for compliance environments.

Interactive Omega Plus setup wizard for coding CLI tools.

MeshCode + MeshGuard. An agentic coding CLI AND a transparent governance proxy in front of Claude Code / Cursor / Cline — every prompt routed through ModelMesh for DLP, audit, budgets, and intent-aware routing.

Teams Developer CLI for managing Microsoft Teams apps

Официальное расширение Brand Map, которое запускает небольшой исполнитель AI-команд как Temporal-процесс `run-harness`.

Shared agent OS infrastructure for Sentinel AI Solutions' product family (Archer, FrankiAI, future agents).

Trail Layer 1: post-hoc capture from Claude Code transcripts. TypeScript port of py-reference/cli/trail.py.

<div align="center">

Pipeline Codex-native para equipes que desenvolvem com IA

 

Local Actor Kernel for Pi

Give vision-blind coding models eyes (DeepSeek + Qwen3.6, any agent). See cost per task, not per model. The dynamic harness for your agent fleet — Free for See; Control + Govern paid.